Ok all,
This website of "tests" is talked up by nearly everyone who wants to win a technical argument about whether different capacitors sound different.
My question is: is it valid, from a technical point of view, as a test of things in the audio range? I am not claiming to be an engineer. But I have questions about the following:
1) the voltages and currents seem high for a simulation of an audio amp.
2) he doesn't say what the scale we're looking at is (or did I miss it?).
3) He does not show a schematic of the test circuit, and I found his description of it hard to follow. Can somebody draw it for us?
I'm worried about any test that needs to be exaggerated to show results... honestly, I'm not trying to start another fight here. It seems that this website, along with the Jung and Marsh article is used as the seminal evidence of these differences, and if it's valid, it can't hurt to look closer at the methods used.
This website of "tests" is talked up by nearly everyone who wants to win a technical argument about whether different capacitors sound different.
My question is: is it valid, from a technical point of view, as a test of things in the audio range? I am not claiming to be an engineer. But I have questions about the following:
1) the voltages and currents seem high for a simulation of an audio amp.
2) he doesn't say what the scale we're looking at is (or did I miss it?).
3) He does not show a schematic of the test circuit, and I found his description of it hard to follow. Can somebody draw it for us?
dorkus said:
I'm worried about any test that needs to be exaggerated to show results... honestly, I'm not trying to start another fight here. It seems that this website, along with the Jung and Marsh article is used as the seminal evidence of these differences, and if it's valid, it can't hurt to look closer at the methods used.
Since this thread deals with Steve Benchs test, perhaps we
should have a link to this test also:
http://members.aol.com/sbench102//cap.html
(Edit: Seems Joel posted the link while I was writing this.)
I agree with you Steve that this test raises a number of
questions. It is unclear exactly what he measures and how he
measures it. He talks about "sampling" the voltage, for instance.
Does he simply mean measuring the voltage, ie. feeding it to
one channel of the scope, or does he indeed mean sampling?
In the latter case, how, and why is this important? He says he
integrates the current, but not how? If he does it the analogue
way he needs a capacitor to do the integration, raising the
question how this capacitor may affect the test. Controversely,
if he samples and integrates numerically, there are other
potential sources of errors which he does not say anything about.
One also wonders if all capacitors had the same value, and
what value this was? For the test to be meaningful, I guess one
has to use components with the same capacitance, but this
will lead to the use of somewhat extreme values for certain
types of capacitors, which may be another source of error.
As for the exaggeration, I think that is reasonable, since the
point was to show the results on the scope. one must then,
however, estimate how to scale the results to reasonable values,
since we seem not to deal with a linear function.
should have a link to this test also:
http://members.aol.com/sbench102//cap.html
(Edit: Seems Joel posted the link while I was writing this.)
I agree with you Steve that this test raises a number of
questions. It is unclear exactly what he measures and how he
measures it. He talks about "sampling" the voltage, for instance.
Does he simply mean measuring the voltage, ie. feeding it to
one channel of the scope, or does he indeed mean sampling?
In the latter case, how, and why is this important? He says he
integrates the current, but not how? If he does it the analogue
way he needs a capacitor to do the integration, raising the
question how this capacitor may affect the test. Controversely,
if he samples and integrates numerically, there are other
potential sources of errors which he does not say anything about.
One also wonders if all capacitors had the same value, and
what value this was? For the test to be meaningful, I guess one
has to use components with the same capacitance, but this
will lead to the use of somewhat extreme values for certain
types of capacitors, which may be another source of error.
As for the exaggeration, I think that is reasonable, since the
point was to show the results on the scope. one must then,
however, estimate how to scale the results to reasonable values,
since we seem not to deal with a linear function.
STEVE'S TEST
Hello,
Christer,here is a partial answer to your QQ:
"I measured several different types of parts, and captured the results (simply by training a camera at the scope). The value of each of the capacitors was constant, 0.1uF. The signal level was held constant at about 70 volts RMS at 600 Hz across the capacitors. (for about 26mA signal current). This is probably more than you would normally expect, and serves to show the results better. A number of capacitor types were used in this experiment. The first series of curves show paper and oil, polycarbonate film, polyester film, polystyrene film, polypropylene, 100 v olt and 1000 volt ceramic and silver mica."
Cheers,
Hello,
Christer,here is a partial answer to your QQ:
"I measured several different types of parts, and captured the results (simply by training a camera at the scope). The value of each of the capacitors was constant, 0.1uF. The signal level was held constant at about 70 volts RMS at 600 Hz across the capacitors. (for about 26mA signal current). This is probably more than you would normally expect, and serves to show the results better. A number of capacitor types were used in this experiment. The first series of curves show paper and oil, polycarbonate film, polyester film, polystyrene film, polypropylene, 100 v olt and 1000 volt ceramic and silver mica."
Cheers,
Frank,
You're right, I missed that he actually did specify the capacitance.
Still, there are a lot of things that are unclear to me and that
should be addressed, by Bench or someone else repeating
the test. Note that I do not claim the results to be wrong, just
that "proof" itself is wanting. There clearly is something he
has captured on the scope, I just want to know more precisely
what it is.
You're right, I missed that he actually did specify the capacitance.
Still, there are a lot of things that are unclear to me and that
should be addressed, by Bench or someone else repeating
the test. Note that I do not claim the results to be wrong, just
that "proof" itself is wanting. There clearly is something he
has captured on the scope, I just want to know more precisely
what it is.
CAP TEST
Hi,
Well,I don't think S.B. claims it is a scientific test.
When it was first published most people agreed that the reuslts correlated with what was expected.
At the bottom of the page there is some explanation.
Btw,I don't think he had numerical (digital) sampling in mind.
In American English "to sample" also means use somtehing as a sample.
"The "current monitor" resistor used for all curves was 100 ohms. The "integrator" was passive, using a 100k and 0.047, 0.1, or 0.47 uF capacitors depending on the frequency and level. The "X" input to the scope was fed with an additional series capacitor and adjusted for exactly "90 degrees" phase shift (typically ran 0.01u into the 10 meg scope probe resistance) to reduce systemic errors. The generator is absolutely isolated, being the secondary of a transformer capable of 300 volts into 1k from 15Hz to 1kHz. The bias voltage was added "in series" with this from an HP regulated and isolated supply. The junction of the current monitor resistor and the capacitor under test serves as the scope and system "reference" point. "
Ciao,
Hi,
Well,I don't think S.B. claims it is a scientific test.
When it was first published most people agreed that the reuslts correlated with what was expected.
At the bottom of the page there is some explanation.
Btw,I don't think he had numerical (digital) sampling in mind.
In American English "to sample" also means use somtehing as a sample.
"The "current monitor" resistor used for all curves was 100 ohms. The "integrator" was passive, using a 100k and 0.047, 0.1, or 0.47 uF capacitors depending on the frequency and level. The "X" input to the scope was fed with an additional series capacitor and adjusted for exactly "90 degrees" phase shift (typically ran 0.01u into the 10 meg scope probe resistance) to reduce systemic errors. The generator is absolutely isolated, being the secondary of a transformer capable of 300 volts into 1k from 15Hz to 1kHz. The bias voltage was added "in series" with this from an HP regulated and isolated supply. The junction of the current monitor resistor and the capacitor under test serves as the scope and system "reference" point. "
Ciao,
Re: CAP TEST
No he doesn't, at least not in the text linked to, but some people
both here and at audioasylum uses the results in debates as if
it were a scientific truth. BTW, as far as I remember there was
quite some discussion on audioasylum about this test and its
validity, so I don't think everbody agrees to the correlation.
Where did you find this info? It certainly adds a lot of valuable
info and should have been published along with the test.
fdegrove said:
No he doesn't, at least not in the text linked to, but some people
both here and at audioasylum uses the results in debates as if
it were a scientific truth. BTW, as far as I remember there was
quite some discussion on audioasylum about this test and its
validity, so I don't think everbody agrees to the correlation.
Where did you find this info? It certainly adds a lot of valuable
info and should have been published along with the test.
Seems we were all right. The info on the test setup was
nowhere to read on the page I linked to. However, after your
claims to the contrary I found that Joel and I have obviously
linked to two different version of the page, in my case one
without this info. I just copied the link from the other thread
where it was brought up, so Joel must have found his link
somewhere else.
Furthermore, don't get me wrong, I think both the test and
its claimed results are very interesting, especially since it seems
to correlate to what people think of the various capacitors
sonically. However, the test is no proof in any reasonable sense,
but an interesting starting point for those who feel tempted
to do more rigorous tests. When I think more about it, I wonder
if this test is anything new, or just new to audio? I think what
the test actually aims to measure is the dielectric absorbtion,
which also what audio people usually claims as the explanation
for the sonic results.
This phenomenon is well-known in electronics, however, so
there should most likely be more extensive tests done
somewhere in the literature, or at least by manufacturers of
capacitors.
nowhere to read on the page I linked to. However, after your
claims to the contrary I found that Joel and I have obviously
linked to two different version of the page, in my case one
without this info. I just copied the link from the other thread
where it was brought up, so Joel must have found his link
somewhere else.
Furthermore, don't get me wrong, I think both the test and
its claimed results are very interesting, especially since it seems
to correlate to what people think of the various capacitors
sonically. However, the test is no proof in any reasonable sense,
but an interesting starting point for those who feel tempted
to do more rigorous tests. When I think more about it, I wonder
if this test is anything new, or just new to audio? I think what
the test actually aims to measure is the dielectric absorbtion,
which also what audio people usually claims as the explanation
for the sonic results.
This phenomenon is well-known in electronics, however, so
there should most likely be more extensive tests done
somewhere in the literature, or at least by manufacturers of
capacitors.
SB tests
Re: yes!
quote:
--------------------------------------------------------------------------------
Originally posted by dorkus
http://members.aol.com/sbench102/caps.html
[snip]and of course the non-film types look completely different altogether.
--------------------------------------------------------------------------------
At any rate is shows that if you put 70VRMS, which is some 200V Pk-Pk on a 100V (or lower, logic bypass) capacitor that the cap starts to break down. Very subtle indeed.
Jan Didden
Re: yes!
quote:
--------------------------------------------------------------------------------
Originally posted by dorkus
http://members.aol.com/sbench102/caps.html
[snip]and of course the non-film types look completely different altogether.
--------------------------------------------------------------------------------
At any rate is shows that if you put 70VRMS, which is some 200V Pk-Pk on a 100V (or lower, logic bypass) capacitor that the cap starts to break down. Very subtle indeed.
Jan Didden
CAP TESTS
Hi,
No,absolutely not.
Similar tests like that were actually carried out by the French and published in issues of the "L'audiophile" magazine.
And that was actually in the seventies and eighties.
Yes you are correct,SB shows the influence of dielectric material at audio frequencies.
All in all an interesting test since it clearly shows that the superior materials actually also measure better.
And this then "proves" what listeners experience when upgrading to a superior capacitor.
Cheers,
Hi,
No,absolutely not.
Similar tests like that were actually carried out by the French and published in issues of the "L'audiophile" magazine.
And that was actually in the seventies and eighties.
Yes you are correct,SB shows the influence of dielectric material at audio frequencies.
All in all an interesting test since it clearly shows that the superior materials actually also measure better.
And this then "proves" what listeners experience when upgrading to a superior capacitor.
Cheers,
Well, but SB then actually doesn't show anything new at all,
I think. The different dielectric properties of various capacitors
is well known in electronics, so he just demonstrates what is
already known, or...? As for the sonic results, we might say
there is some kind of correlation, although not in the statistic
sense. Some people may wish to use the word proof carelessly,
I don't.
I think. The different dielectric properties of various capacitors
is well known in electronics, so he just demonstrates what is
already known, or...? As for the sonic results, we might say
there is some kind of correlation, although not in the statistic
sense. Some people may wish to use the word proof carelessly,
I don't.
CAP TESTS
Joel,
If you're mad about that then I really don't understand what this thread is about.
It is however what the test by S.B. was about.
I did not even express MY opinion but a fact.
If you care to dig into industrial manuals or product datasheets you will frequently find graphs showing freqency responses of various caps.
You notice the differences in frequency response versus dielitric absortion etc.
Cheers, (looky,I keep smiling)
Joel,
If you're mad about that then I really don't understand what this thread is about.
It is however what the test by S.B. was about.
I did not even express MY opinion but a fact.
If you care to dig into industrial manuals or product datasheets you will frequently find graphs showing freqency responses of various caps.
You notice the differences in frequency response versus dielitric absortion etc.
Cheers,
(looky,I keep smiling)
I was only "mad" because you made a statement like this:
And this then "proves" what listeners experience when upgrading to a superior capacitor.
The test may show that, or it may not. That is what I'd like proven.
Let me give you an example: there are measuring devices that will clearly show you in New York when there is an earthquake in California. But obviously no one in NY said "hey I feel an earthquake!!!!".
Steve (and the manuals) are showing problems with the dieletrics under certain conditions? Ok. But does his measurement show that the non-linearities deviate by what should be an audible amount? He doesn't say, and neither does anyone else.
And this then "proves" what listeners experience when upgrading to a superior capacitor.
The test may show that, or it may not. That is what I'd like proven.
Let me give you an example: there are measuring devices that will clearly show you in New York when there is an earthquake in California. But obviously no one in NY said "hey I feel an earthquake!!!!".
Steve (and the manuals) are showing problems with the dieletrics under certain conditions? Ok. But does his measurement show that the non-linearities deviate by what should be an audible amount? He doesn't say, and neither does anyone else.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.